This repository contains the core router for Python functions (FaaS). The framework is based on Flask and is designed to be used with containers.

task build

This will create base image with the name function/python. The tag version has the following format: <image_version>-<python_version>. For example, 1.0.0-3.11.

The list of builded Python versions can be found in the Taskfile.yml file.

You will need to create a new Dockerfile that inherits from the base image. Then you can add your function code and install the dependencies. See examples.

Function listen on port 8080 by default. You can change it by setting the PORT environment variable.

Function name is handler located in the function.py file by default. You can change it by setting the FUNCTION_HANDLER environment variable. The format is <module_name>:<function_name>.

Function list on path / by default. You can change it by setting the FUNCTION_PATH environment variable.

Function listen to method GET and POST by default. You can change it by setting the FUNCTION_METHODS environment variable (Comma separated list).

Function listen on /healthz path for health check.

First, you need to build the image. See examples if you don't know where to start.

With Docker

docker run -p 8080:8080 --rm <image_name>

With Kubernetes

kubectl run <pod_name> --image=<image_name> --port=8080
kubectl port-forward <pod_name> 8080:8080

curl -X POST http://localhost:8080/

This project uses Taskfile to automate the development process. You can find the list of available commands in the Taskfile.yml file.

Source code is located in the app directory. There are also Python requirements in the requirements.txt file. The scripts folder holds runtime script like the image entrypoint. Finally, the Dockerfile is used to build the base image.

The webserver used is Gunicorn.

The base application framework is Flask.